# Infinitely many solutions via critical points for a fractional p-Laplacian equation with perturbations

## Abstract

In this paper, we use variant fountain theorems to study the existence of infinitely many solutions for the fractional p-Laplacian equation

$$(-\Delta )_{p}^{\alpha }u+\lambda V(x) \vert u \vert ^{p-2}u=f(x,u)-\mu g(x) \vert u \vert ^{q-2}u,\quad x\in \mathbb{R}^{N},$$

where $$\lambda,\mu$$ are two positive parameters, $$N,p\ge 2$$, $$q\in (1,p)$$, $$\alpha \in (0,1)$$, $$(-\Delta )_{p}^{\alpha }$$ is the fractional p-Laplacian, and $$V,g,u:\mathbb{R}^{N}\to \mathbb{R}$$, $$f:\mathbb{R}^{N}\times \mathbb{R}\to \mathbb{R}$$.

## Introduction

In this paper we investigate the existence of infinitely many solutions for the fractional p-Laplacian equation

$$(-\Delta )_{p}^{\alpha }u+\lambda V(x) \vert u \vert ^{p-2}u=f(x,u)-\mu g(x) \vert u \vert ^{q-2}u, \quad x\in \mathbb{R}^{N},$$
(1)

where $$\lambda,\mu$$ are two positive parameters, $$N,p\ge 2$$, $$\alpha \in (0,1)$$, $$(-\Delta )_{p}^{\alpha }$$ is the fractional p-Laplacian, and the potential function $$V:\mathbb{R}^{N}\to \mathbb{R}$$ satisfies the following conditions:

1. (V1)

$$V\in C(\mathbb{R}^{N}, \mathbb{R})$$ and $$\inf_{x\in \mathbb{R} ^{N}} {V}(x)\ge V_{0}>0$$, where $$V_{0}$$ is a positive constant.

2. (V2)

There exists $$b>0$$ such that $$\text{meas}\{x\in \mathbb{R}^{N}: {V}(x)\le b\}$$ is finite, where meas denotes the Lebesgue measures.

The functions $$f:\mathbb{R}^{N}\times \mathbb{R}\to \mathbb{R}$$, $$g:\mathbb{R}^{N}\to \mathbb{R}$$ satisfy the conditions:

1. (f1)

$${f}\in C(\mathbb{R}^{N}\times \mathbb{R},\mathbb{R})$$ and $$\lim_{|u|\to 0} \frac{f(x,u)}{|u|^{p-2}u}=0$$ uniformly in $$x\in \mathbb{R}^{N}$$.

2. (f2)

$${F}(x,u)=\int _{0}^{u} {f}(x,s)\,\mathrm{d}s\ge 0$$ and $$\mathscr{{F}}(x,u)= \frac{1}{p}{f}(x,u)u-{F}(x,u)\ge 0$$ for all $$(x,u)\in \mathbb{R}^{N} \times \mathbb{R}$$.

3. (f3)

$$\lim_{|u|\to \infty }\frac{{f}(x,u)u}{|u|^{p}}=+\infty$$ uniformly in $$x\in \mathbb{R}^{N}$$.

4. (f4)

There exist $$d_{1},r_{0}>0$$ and $$\tau > \frac{p_{\alpha }^{*}}{p _{\alpha }^{*}-p}$$ with $$p_{\alpha }^{*}= \bigl\{ \scriptsize{ \begin{array}{l@{\quad}l} \frac{Np}{N-\alpha p} & \text{if }\alpha p< N, \\ \infty &\text{if }\alpha p\ge N, \end{array}}$$ such that

$$\bigl\vert {f}(x,u) \bigr\vert ^{\tau }\le d_{1} \mathscr{{F}}(x,u) \vert u \vert ^{(p-1)\tau }\quad\text{for all } x\in \mathbb{R}^{N} \text{ and } \vert u \vert \ge r_{0}.$$
5. (f5)

$$f(x,-u)=-f(x,u)$$ for all $$(x,u)\in \mathbb{R}^{N}\times \mathbb{R}$$.

6. (g)

$$g\in L^{q'}(\mathbb{R}^{N})$$ and $$g(x)\ge 0$$ $$(\not \equiv 0)$$ for a.e. $$x\in \mathbb{R}^{N}$$, where $$q'\in (\frac{p_{\alpha }^{*}}{p _{\alpha }^{*}-q}, \frac{p}{p-q} ], q\in (1,p)$$.

Fractional systems arise for example in phase transitions, chaos, diffusion, finance, flame propagation, and wave propagation. In , the authors introduced a fractional order modified Duffing system

$$\textstyle\begin{cases} \frac{\mathrm{d}^{q_{1}}x}{\mathrm{d}t^{q_{1}}}=y,\qquad \frac{\mathrm{d}^{q_{2}}y}{\mathrm{d}t ^{q_{2}}}=-x-x^{3}-ay+bz, \\ \frac{\mathrm{d}z}{\mathrm{d}t}=w, \qquad \frac{\mathrm{d}w}{\mathrm{d}t}=-cz-dz^{3}, \end{cases}$$

where $$\frac{\mathrm{d}^{q_{1}}x}{\mathrm{d}t^{q_{1}}},\frac{\mathrm{d}^{q_{2}}y}{\mathrm{d}t ^{q_{2}}}$$ are fractional derivatives, and via phase portraits and bifurcation diagrams, they studied chaotic behaviors for this system; we also refer the reader to the books [2,3,4] and the papers [5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23]. Variational methods and critical point theory were used to study fractional Schrödinger equations in the literature [24,25,26,27,28,29,30,31,32,33,34,35,36,37]; for results on Schrödinger equations, we refer the reader to [38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66]. In [24, 25], Ambrosio and Torres used the mountain pass theorem and a variant of the fountain theorem to obtain the existence of nontrivial solutions for (1) with $$\lambda =1,\mu =0$$, where f is p-superlinear at infinity. In , Tang et al. obtained infinitely many solutions for the following fractional p-Laplacian equations of Schrödinger–Kirchhoff type:

$$\biggl(a+b \iint _{\mathbb{R}^{2N}}\frac{ \vert u(x)-u(y) \vert ^{p}}{ \vert x-y \vert ^{N+p \alpha }}\,\mathrm{d}x\,\mathrm{d}y \biggr)^{p-1}(- \Delta )_{p}^{\alpha }u+ V(x) \vert u \vert ^{p-2}u=f(x,u),$$
(2)

where they used the condition:

(Tang) There exist $$c_{0}>0,r_{0}>0$$, and $$\kappa >\{1, \frac{N}{p \alpha }\}$$ such that

$$\bigl\vert F(x,t) \bigr\vert ^{\kappa }\le c_{0} \vert t \vert ^{p\kappa } \biggl[\frac{1}{p^{2}}f(x,t)t-F(x,t) \biggr], \quad\forall (x,t)\in \mathbb{R}^{N}\times \mathbb{R}, \vert t \vert \ge r_{0},$$

to ensure that the energy functional satisfies the Palais–Smale condition, i.e., (PS) sequence has a convergent subsequence; this condition can also be found in [26, 28, 40,41,42]. There are only a few papers on (1) with a sublinear perturbation. For example, in  the authors used the famous Ambrosetti–Rabinowitz condition:

(AR) There exists $$\mu >p^{2}$$ such that

$$0< \mu F(x,t)\le f(x,t)t, \quad\forall x\in \mathbb{R}^{N}, t\in \mathbb{R} \backslash \{0\},$$

to obtain nontrivial solutions for (2) with a perturbation g ($$g\in L^{\frac{p}{p-1}}(\mathbb{R}^{N})$$). In [30,31,32, 38, 39] similar methods were used to study various Schrödinger equations with perturbations.

Motivated by the above papers, in this paper we use variant fountain theorems to study the existence of nontrivial solutions for the fractional p-Laplacian equation (1). The novelty is two-fold: (i) the condition (Tang) is adopted to ensure that bounded sequences have convergent subsequences, (ii) we consider the influence of parameters and perturbation terms on the existence of solutions.

Now, we state our main result.

### Theorem 1.1

Suppose that (V1)–(V2), (f1)–(f5), and (g) hold. Then, for sufficiently small $$\mu >0$$, there exists $$\varLambda >0$$ such that system (1) possesses infinitely many solutions when $$\lambda \ge \varLambda$$.

### Remark 1.2

Note that (f1), (f2), and (f4) imply that f has subcritical growth. From (f2), (f4), for all $$x\in \mathbb{R}^{N}, |u| \ge r_{0}$$, we find

\begin{aligned} \bigl\vert {f}(x,u) \bigr\vert ^{\tau } & \le d_{1}\mathscr{{F}}(x,u) \vert u \vert ^{(p-1)\tau }=d_{1} \biggl(\frac{1}{p}{f}(x,u)u-{F}(x,u) \biggr) \vert u \vert ^{(p-1)\tau } \\ &\le \frac{d _{1}}{p} \bigl\vert {f}(x,u) \bigr\vert \vert u \vert ^{(p-1)\tau +1}. \end{aligned}

This shows that

$$\bigl\vert {f}(x,u) \bigr\vert ^{\tau -1} \le \frac{d_{1}}{p} \vert u \vert ^{(p-1) \tau +1} \quad\text{and}\quad \bigl\vert {f}(x,u) \bigr\vert \le \sqrt[\tau -1]{\frac{d_{1}}{p}} \vert u \vert ^{\frac{(p-1) \tau +1}{\tau -1}}.$$

Let $$\frac{(p-1)\tau +1}{\tau -1}=s-1$$. Then $$s= \frac{p\tau }{\tau -1}\in (p,p_{\alpha }^{*})$$. On the other hand, from (f1) for all $$\varepsilon >0$$, we have

$$\bigl\vert f(x,u) \bigr\vert \le \varepsilon \vert u \vert ^{p-1} \quad\text{for }x\in \mathbb{R}^{N}, \vert u \vert \le r_{0},$$

and hence, there exists $$c_{\varepsilon }=\sqrt[\tau -1]{ \frac{d_{1}}{p}}>0$$ such that

$$\bigl\vert {f}(x,u) \bigr\vert \le \varepsilon \vert u \vert ^{p-1}+c_{\varepsilon } \vert u \vert ^{s-1},\quad \forall (x,u)\in \mathbb{R}^{N}\times \mathbb{R},$$
(3)

and from $${F}(x,u)=\int _{0}^{u} {f}(x,s)\,\mathrm{d}s$$ we have

$$\bigl\vert {F}(x,u) \bigr\vert \le \frac{\varepsilon }{p} \vert u \vert ^{p}+ \frac{c_{\varepsilon }}{s} \vert u \vert ^{s},\quad \forall (x,u)\in \mathbb{R}^{N} \times \mathbb{R}.$$
(4)

### Remark 1.3

Consider the Ambrosetti–Rabinowitz condition (see [29,30,31,32, 38, 39]):

(AR) There exists $$\theta >p$$ such that

$$0< \theta F(x,u)\le f(x,u)u\quad \text{for all } x\in \mathbb{R}^{N}, u \in \mathbb{R}\backslash \{0\}.$$

Let $$F(x,u)=|\sin x||u|^{p}\ln (1+|u|), \forall x\in \mathbb{R}^{N}, u\in \mathbb{R}$$. Then $$f(x,u)=|\sin x| (p|u|^{p-2}u\ln (1+|u|) + \frac{|u|^{p-1}u}{1+|u|} )$$. Consequently, for all $$x\in \mathbb{R}^{N}$$, we have

$$\theta F(x,u)-f(x,u)u= \vert \sin x \vert (\theta -p) \vert u \vert ^{p}\ln \bigl(1+ \vert u \vert \bigr)- \vert \sin x \vert \frac{ \vert u \vert ^{p+1}}{1+ \vert u \vert }\le 0,$$

and this is impossible for large $$|u|$$. However, this function satisfies conditions (f1)–(f5).

## Preliminaries

We first discuss the space $$W^{\alpha,p}(\mathbb{R}^{N})$$ (for more details, we refer the reader to ). When $$u:\mathbb{R}^{N} \to \mathbb{R}$$ is a measurable function, we define the Gagliardo seminorm as follows:

$$[u]_{\alpha,p}:= \biggl[ \int _{\mathbb{R}^{N}} \int _{\mathbb{R}^{N}} \frac{ \vert u(x)-u(y) \vert ^{p}}{ \vert x-y \vert ^{N+ \alpha p}} \,\mathrm{d}x \,\mathrm{d}y \biggr]^{\frac{1}{p}},\quad p\ge 2.$$

Now, the fractional Sobolev space is given by

$$W^{\alpha,p}\bigl(\mathbb{R}^{N}\bigr):=\bigl\{ u\in L^{p}\bigl(\mathbb{R}^{N}\bigr): u \text{ is measurable and } [u]_{\alpha,p}< \infty \bigr\} ,$$

with the norm

$$\Vert u \Vert _{\alpha,p} = \bigl( [u]_{\alpha,p}^{p}+ \Vert u \Vert _{p}^{p} \bigr) ^{\frac{1}{p}},$$

where $$\|u\|_{p}$$ is the norm for the usual Lebesgue space $$L^{p}( \mathbb{R}^{N})$$, denoted by

$$\Vert u \Vert _{p}= \biggl( \int _{\mathbb{R}^{N}} \bigl\vert u(x) \bigr\vert ^{p} \,\mathrm{d}x \biggr)^{ \frac{1}{p}}.$$

For the potential function V, we consider the following fractional Sobolev space:

$$E:= \biggl\{ u\in W^{\alpha,p}\bigl(\mathbb{R}^{N}\bigr): \int _{\mathbb{R}^{N}} V(x) \bigl\vert u(x) \bigr\vert ^{p} \,\mathrm{d}x< \infty \biggr\} ,$$

with the norm

$$\Vert u \Vert _{E}:= \biggl( [u]_{\alpha,p}^{p} + \int _{\mathbb{R}^{N}} V(x) \bigl\vert u(x) \bigr\vert ^{p} \,\mathrm{d}x \biggr)^{\frac{1}{p}}.$$

Note that the parameter λ can be chosen large enough, so this norm can be replaced by

$$\Vert u \Vert := \biggl( [u]_{\alpha,p}^{p} + \int _{\mathbb{R}^{N}} \lambda V(x) \bigl\vert u(x) \bigr\vert ^{p} \,\mathrm{d}x \biggr)^{\frac{1}{p}}.$$

In summary, throughout our paper we use the space $$(E,\|\cdot \|)$$.

### Lemma 2.1

(see [67, Theorem 6.5] and [25, Lemma 2.1])

The embedding $$E\hookrightarrow L^{t}( \mathbb{R}^{N})$$ is continuous if $$t\in [p,p_{\alpha }^{*}]$$ and compact if $$t\in [p,p_{\alpha }^{*})$$.

Hence, there exists $$C_{t}>0$$ such that

$$\Vert u \Vert _{t}\le C_{t} \Vert u \Vert , \quad\forall t\in \bigl[p,p_{\alpha }^{*}\bigr].$$
(5)

Let X be a reflexive and separable Banach space and $$X^{*}$$ be its dual space. Then there are (see [68, Sect. 17]) $$\{\phi _{n}\} _{n\in \mathbb{N}}\subset X$$ and $$\{\phi _{n}^{*}\}_{n\in \mathbb{N}} \subset X^{*}$$ such that $$X=\overline{\text{span}\{\phi _{n}:n\in \mathbb{N}\}}$$, $$X^{*}=\overline{\text{span}\{\phi _{n}^{*}:n\in \mathbb{N}\}}$$, and $$\langle \phi _{n},\phi _{m}\rangle = \bigl\{ \scriptsize{ \begin{array}{l@{\quad}l} 1,& n=m, \\ 0,& n\neq m. \end{array}}$$ For $$k=1,2,\ldots$$ , let $$Y_{k}=\text{span}\{\phi _{1},\ldots,\phi _{k}\}$$ and $$Z_{k}=\overline{\text{span}\{\phi _{k},\phi _{k+1},\ldots\}}$$.

### Lemma 2.2

(see )

Let X be a Banach space, and $$X=\overline{\bigoplus_{j\in \mathbb{N}}X_{j}}$$ with $$\dim X_{j}<\infty$$ for any $$j\in \mathbb{N}$$. Set $$Y_{k}=\bigoplus_{j=0}^{k} X_{j}, Z_{k}=\overline{ \bigoplus_{j=k+1}^{\infty }X_{j}}$$. Consider the following $$C^{1}$$ functional $$\varPhi _{\lambda }: X\to \mathbb{R}$$ defined by

$$\varPhi _{\lambda }(u)=A(u)-\lambda B(u),\quad \lambda \in [1,2].$$

Suppose that

1. (Z1)

$$\varPhi _{\lambda }$$ maps bounded sets to bounded sets uniformly for $$\lambda \in [1,2]$$. Furthermore, $$\varPhi _{\lambda }(-u)=\varPhi _{\lambda }(u)$$ for $$(\lambda,u)\in [1,2]\times X$$;

2. (Z2)

$$B(u)\ge 0$$; $$B(u)\to \infty$$ as $$\|u\|\to \infty$$ on any finite dimensional subspace of X;

3. (Z3)

There exist $$\rho _{k}>r_{k}>0$$ such that $$a_{k}(\lambda )= \inf_{u\in Z_{k},\|u\|=\rho _{k}}\varPhi _{\lambda }(u)\ge 0>b_{k}(\lambda )=\max_{u\in Y_{k},\|u\|=r_{k}}\varPhi _{\lambda }(u)$$ for $$\lambda \in [1,2]$$, $$d_{k}(\lambda )=\inf_{u\in Z_{k},\|u\|\le \rho _{k}} \varPhi _{\lambda }(u)\to 0$$ as $$k\to \infty$$, uniformly for $$\lambda \in [1,2]$$.

Then there exist $$\lambda _{n}\to 1$$, $$u(\lambda _{n})\in Y_{n}$$ such that $$\varPhi '_{\lambda _{n}}|_{Y_{n}}(u(\lambda _{n}))=0$$, $$\varPhi _{\lambda _{n}}(u( \lambda _{n}))\to c_{k}\in [d_{k}(2),b_{k}(1)]$$ as $$n\to \infty$$. In particular, if $$\{u(\lambda _{n})\}$$ has a convergent subsequence for every k, then $$\varPhi _{1}$$ has infinitely many nontrivial critical points $$\{u_{k}\}\subset X\backslash \{0\}$$ satisfying $$\varPhi _{1}(u _{k})\to 0^{-}$$ as $$k\to \infty$$.

## Main results

Now, we can define the energy functional J on E as follows:

$$J(u)=\frac{1}{p} \Vert u \Vert ^{p}- \int _{\mathbb{R}^{N}} F(x,u)\,\mathrm{d}x+\frac{ \mu }{q} \int _{\mathbb{R}^{N}} g(x) \vert u \vert ^{q}\,\mathrm{d}x \quad\text{for } x\in \mathbb{R}^{N}, u\in E.$$
(6)

From (4), (V1)–(V2), and (g) we have that J is well defined and of class $$C^{1}$$. Moreover,

\begin{aligned} \bigl\langle J'(u),\varphi \bigr\rangle ={}& \int _{\mathbb{R}^{N}} \int _{\mathbb{R} ^{N}} \frac{ \vert u(x)-u(y) \vert ^{p-2}(u(x)-u(y))(\varphi (x)-\varphi (y))}{ \vert x-y \vert ^{N+ \alpha p}}\,\mathrm{d}x\,\mathrm{d}y \\ &{} + \int _{\mathbb{R}^{N}} \lambda V(x) \vert u \vert ^{p-2}u \varphi \,\mathrm{d}x \\ &{} - \int _{\mathbb{R}^{N}} f(x,u)\varphi \,\mathrm{d}x+\mu \int _{\mathbb{R}^{N}} g(x) \vert u \vert ^{q-2}u\varphi \,\mathrm{d}x \quad\text{for } x\in \mathbb{R}^{N}, u, \varphi \in E. \end{aligned}
(7)

From the definition of $$J'$$, we see that the critical points of J are weak solutions for (1). From , we know that the space E can be decomposed as X in Lemma 2.2, so we can consider the family of functionals $$J_{\nu }: E \to \mathbb{R}$$ defined by

$$J_{\nu }(u)= \frac{1}{p} \Vert u \Vert ^{p}+ \frac{\mu }{q} \int _{\mathbb{R}^{N}} g(x) \vert u \vert ^{q}\,\mathrm{d}x- \nu \int _{\mathbb{R}^{N}} F(x,u)\,\mathrm{d}x := A(u)- \nu B(u) \quad\text{for } \nu \in [1,2].$$

Then $$B(u)\ge 0$$ for $$u\in E$$, and $$J_{\nu }(-u)=J_{\nu }(u)$$ for $$(\nu,u)\in [1,2]\times E$$. Also, it is easy to see that $$J_{\nu }$$ maps bounded sets to bounded sets uniformly on $$\nu \in [1,2]$$.

### Lemma 3.1

Suppose that the assumptions of Theorem 1.1 hold. Then $$B(u)\to \infty$$ as $$\|u\|\to \infty$$ on any finite dimensional subspace of E.

### Proof

For any finite dimensional subspace $$\widetilde{E} \subset E$$, there exists $$\varepsilon _{1}>0$$ such that

$$\text{meas}\bigl\{ x\in \mathbb{R}^{N}: \bigl\vert u(x) \bigr\vert ^{p}\ge \varepsilon _{1} \Vert u \Vert ^{p}\bigr\} \ge \varepsilon _{1}, \quad\forall u\in \widetilde{E}\backslash \{0 \}.$$
(8)

If (8) is not true, then for all $$n\in \mathbb{N}$$, there exists $$u_{n}\in \widetilde{E}\backslash \{0\}$$ such that

$$\text{meas} \biggl\{ x\in \mathbb{R}^{N}: \bigl\vert u_{n}(x) \bigr\vert ^{p}\ge \frac{1}{n} \Vert u_{n} \Vert ^{p} \biggr\} < \frac{1}{n}.$$

Define $$v_{n}(x)=\frac{u_{n}(x)}{\|u_{n}\|}\in \widetilde{E}\backslash \{0\}$$, then for all $$n\in \mathbb{N}$$, $$\|v_{n}\|=1$$, and we obtain

$$\text{meas} \biggl\{ x\in \mathbb{R}^{N}: \bigl\vert v_{n}(x) \bigr\vert ^{p}\ge \frac{1}{n} \biggr\} < \frac{1}{n}.$$
(9)

Since $$\dim \widetilde{E}<\infty$$, passing to a subsequence if necessary, we may assume that $$v_{n}\to v_{0}$$ in . Moreover, $$\|v_{0}\|=1$$. From the equivalence of all norms on the finite dimensional space , we have

$$\int _{\mathbb{R}^{N}} \vert v_{n}-v_{0} \vert ^{p} \,\mathrm{d}x \to 0, \quad\text{as } n \to \infty.$$
(10)

Thus, there exist $$\xi _{1},\xi _{2}>0$$ such that

$$\text{meas} \bigl\{ x\in \mathbb{R}^{N}: \bigl\vert v_{0}(x) \bigr\vert ^{p}\ge \xi _{1} \bigr\} \ge \xi _{2}.$$
(11)

If not, for all $$n\in \mathbb{N}$$, we obtain

$$\text{meas} \biggl\{ x\in \mathbb{R}^{N}: \bigl\vert v_{0}(x) \bigr\vert ^{p}\ge \frac{ 1 }{n} \biggr\} =0.$$

This implies that

$$0\le \int _{\mathbb{R}^{N}} \bigl\vert v_{0}(x) \bigr\vert ^{2p} \,\mathrm{d}x< \frac{1}{n} \Vert v_{0} \Vert _{p}^{p} \le \frac{C_{p}^{p}}{n} \Vert v_{0} \Vert ^{p}= \frac{C_{p}^{p}}{n} \to 0, \quad\text{as } n\to \infty, \text{ for some } C_{p}>0.$$

Hence, $$v_{0}=0$$, contradicting $$\|v_{0}\|=1$$, and then (11) holds.

Now let

\begin{aligned} &\varOmega _{0}= \bigl\{ x\in \mathbb{R}^{N}: \bigl\vert v_{0}(x) \bigr\vert ^{p} \ge \xi _{1} \bigr\} ,\qquad \varOmega _{n}= \biggl\{ x\in \mathbb{R}^{N}: \bigl\vert v_{n}(x) \bigr\vert ^{p} < \frac{1}{n} \biggr\} \quad\text{and}\\ &\varOmega _{n}^{c}=\mathbb{R}^{N} \backslash \varOmega _{n}= \biggl\{ x\in \mathbb{R}^{N}: \bigl\vert v_{n}(x) \bigr\vert ^{p} \ge \frac{1}{n} \biggr\} . \end{aligned}

From (9) and (11), we have

$$\text{meas}(\varOmega _{n}\cap \varOmega _{0})\ge \text{meas}(\varOmega _{0})- \text{meas}\bigl(\varOmega _{n}^{c}\cap \varOmega _{0}\bigr)\ge \xi _{2}-\frac{1}{n}, \quad\forall n\in \mathbb{N}.$$

For n large enough (for example, taking n such that $$\xi _{2}- \frac{1}{n}\ge \frac{1}{2}\xi _{2},\frac{1}{2^{p-1}} \xi _{1} - \frac{1}{n}\ge \frac{1}{2^{p}} \xi _{1}$$), using the inequality $$|v_{n}|^{p}=|v_{n}-v_{0}+v_{0}|^{p} \le 2^{p-1} |v_{n}-v_{0}|^{p} +2^{p-1} |v_{0}|^{p}$$, for $$p\ge 2$$, we have

\begin{aligned} \int _{\mathbb{R}^{N}} \vert v_{n}-v_{0} \vert ^{p} \,\mathrm{d}x & \ge \int _{\varOmega _{n}\cap \varOmega _{0}} \vert v_{n}-v_{0} \vert ^{p} \,\mathrm{d}x \\ &\ge \frac{1}{2^{p-1}} \int _{\varOmega _{n}\cap \varOmega _{0}} \bigl\vert v_{0}(x) \bigr\vert ^{p} \,\mathrm{d}x- \int _{\varOmega _{n}\cap \varOmega _{0}} \bigl\vert v_{n}(x) \bigr\vert ^{p} \,\mathrm{d}x \\ & \ge \biggl(\frac{1}{2^{p-1}} \xi _{1} - \frac{1}{n} \biggr) \text{meas}( \varOmega _{n}\cap \varOmega _{0}) \\ &\ge \biggl( \frac{1}{2^{p-1}} \xi _{1} - \frac{1}{n} \biggr) \biggl( \xi _{2}-\frac{1}{n} \biggr) \ge \frac{\xi _{1}\xi _{2}}{2^{p+1}} >0. \end{aligned}

This contradicts (10). As a result, (8) holds. For $$\varepsilon _{1}$$ in (8), let

$$\varOmega _{u}=\bigl\{ x\in \mathbb{R}^{N}: \bigl\vert u(x) \bigr\vert ^{p}\ge \varepsilon _{1} \Vert u \Vert ^{p}\bigr\} ,\quad \forall u\in \widetilde{E}\backslash \{0\}.$$

Then we have $$\text{meas}(\varOmega _{u})\ge \varepsilon _{1}$$. On the other hand, from L’Hospital rule and (f3) we have

$$\lim_{ \vert u \vert \to \infty }\frac{{F}(x,u)}{ \vert u \vert ^{p}}=+\infty\quad \text{uniformly in } x \in \mathbb{R}^{N}.$$

Hence, there exists sufficiently large $$d_{2}>0$$ such that

$$F(x,u)\ge d_{2} \vert u \vert ^{p} \quad\text{ for }x\in \mathbb{R}^{N}, \vert u \vert >r_{1}, \text{ for some } r_{1}>0.$$

From (4) with $$s\in (p,p_{\alpha }^{*})$$, we have

$$F(x,u)\le \vert u \vert ^{p} \biggl(\frac{c_{1}}{p}+ \frac{c_{2}}{s} \vert u \vert ^{s-p} \biggr) \le \biggl( \frac{c_{1}}{p}+\frac{c_{2}}{s} r_{1}^{s-p} \biggr) \vert u \vert ^{p} \quad\text{for }x\in \mathbb{R}^{N}, \vert u \vert \le r_{1}.$$

As a result, there exists $$d_{3}\in (0,d_{2})$$ such that

$$F(x,u)\ge (d_{2}-d_{3}) \vert u \vert ^{p} \quad\text{for } x\in \mathbb{R}^{N}.$$
(12)

This, together with (8), implies that

\begin{aligned} B(u) &= \int _{\mathbb{R}^{N}} F(x,u)\,\mathrm{d}x\ge (d_{2}-d_{3}) \int _{\mathbb{R}^{N}} \bigl\vert u(x) \bigr\vert ^{p} \,\mathrm{d}x \ge (d_{2}-d_{3}) \int _{\varOmega _{u}} \bigl\vert u(x) \bigr\vert ^{p} \,\mathrm{d}x \\ & \ge \varepsilon _{1} (d_{2}-d _{3}) \Vert u \Vert ^{p} \text{meas} (\varOmega _{u})\ge \varepsilon ^{2}_{1} (d_{2}-d _{3}) \Vert u \Vert ^{p}. \end{aligned}
(13)

Thus $$B(u)\to \infty$$ as $$\|u\|\to \infty$$ on any finite dimensional subspace of E. This completes the proof. □

### Lemma 3.2

Suppose that the assumptions of Theorem 1.1 hold. Then there exists a sequence $$\rho _{k}\to 0^{+}$$ as $$k\to \infty$$ such that

$$a_{k}(\nu )=\inf_{u\in Z_{k},\|u\|=\rho _{k}} J_{\nu }(u)\ge 0,$$
(14)

and

$$d_{k}(\nu )=\inf_{u\in Z_{k},\|u\|\le \rho _{k}} J_{\nu }(u)\to 0, \quad\textit{as } k\to \infty, \textit{ uniformly for } \nu \in [1,2],$$
(15)

where $$Z_{k}=\overline{\bigoplus_{j=k}^{\infty }X_{j}}$$ for all $$k\in \mathbb{N}$$.

### Proof

Let $$\beta _{s}(k)=\sup_{u\in Z_{k},\|u\|=1}\|u\|_{s}$$ with $$s\in (p,p_{\alpha }^{*})$$. Then from Lemma 3.8 of  and Lemma 2.1, we have $$\beta _{s}(k)\to 0, k\to \infty$$. Now, for $$u\in Z_{k}$$, from (4), (5), we obtain

\begin{aligned} J_{\nu }(u) &=\frac{1}{p} \Vert u \Vert ^{p}-\nu \int _{\mathbb{R}^{N}} F(x,u)\,\mathrm{d}x+\frac{ \mu }{q} \int _{\mathbb{R}^{N}} g(x) \vert u \vert ^{q}\,\mathrm{d}x \\ & \ge \frac{1}{p} \Vert u \Vert ^{p}-2 \int _{\mathbb{R}^{N}} F(x,u)\,\mathrm{d}x \\ & \ge \frac{1}{p} \Vert u \Vert ^{p} - \frac{2\varepsilon }{p} \Vert u \Vert _{p}^{p} - \frac{2c_{\varepsilon }}{s} \Vert u \Vert _{s}^{s} \\ & \ge \frac{1}{p} \Vert u \Vert ^{p} - \frac{2\varepsilon }{p} C_{p}^{p} \Vert u \Vert ^{p} - \frac{2c_{\varepsilon }}{s}\beta _{s}^{s} (k) \Vert u \Vert ^{s}. \end{aligned}

Let $$\|u\|=\rho _{k}=\beta _{s}(k), u\in Z_{k}$$, note that $$\beta _{s}(k)$$ can be chosen arbitrarily small when k is large, and if $$\varepsilon =\frac{p}{2C_{p}^{p}} [\frac{1}{p}-\frac{3c_{\varepsilon }}{s} \beta _{s}^{s} (k) ]$$, we have

$$J_{\nu }(u)\ge \biggl[ \frac{1}{p} - \frac{2\varepsilon }{p} C_{p} ^{p} -\frac{2c_{\varepsilon }}{s}\beta _{s}^{s} (k) \biggr] \Vert u \Vert ^{s} = \frac{c_{\varepsilon }}{s}\beta _{s}^{s+1} (k)\ge 0\quad \text{for large } k.$$

On the other hand, for any $$u\in Z_{k}$$ with $$\|u\|\le \rho _{k}$$, we have

$$J_{\nu }(u)\ge -\frac{2c_{\varepsilon }}{s}\beta _{s}^{s} (k) \Vert u \Vert ^{s}.$$

Hence,

$$0\ge \inf_{u\in Z_{k}, \Vert u \Vert \le \rho _{k}} J_{\nu }(u)\ge -\frac{2c_{ \varepsilon }}{s} \beta _{s}^{s} (k) \Vert u \Vert ^{s}.$$

Since, $$\rho _{k}\to 0$$ as $$k\to \infty$$, we have

$$d_{k}(\nu )=\inf_{u\in Z_{k},\|u\|\le \rho _{k}} J_{\nu }(u) \to 0,\quad \text{as } k\to \infty \text{ uniformly for } \nu \in [1,2].$$

This completes the proof. □

### Lemma 3.3

Suppose that all the assumptions of Theorem 1.1 hold (and μ is sufficiently small). For the sequence $$\{\rho _{k}\} _{k\in \mathbb{N}}$$ in Lemma 3.2, there exists $$r_{k}\in (0,\rho _{k})$$ for $$k\in \mathbb{N}$$ such that

$$b_{k}(\nu )=\max_{u\in Y_{k},\|u\|=r_{k}} J_{\nu }(u)< 0 \quad\textit{for } \nu \in [1,2],$$
(16)

where $$Y_{k}=\overline{\bigoplus_{j=1}^{k} X_{j}}$$ for $$k\in \mathbb{N}$$.

### Proof

For $$u\in Y_{k}$$, from (13) and (5) we have

\begin{aligned} J_{\nu }(u) &=\frac{1}{p} \Vert u \Vert ^{p}-\nu \int _{\mathbb{R}^{N}} F(x,u)\,\mathrm{d}x+\frac{ \mu }{q} \int _{\mathbb{R}^{N}} g(x) \vert u \vert ^{q}\,\mathrm{d}x \\ & \le \frac{1}{p} \Vert u \Vert ^{p}- \int _{\mathbb{R}^{N}} F(x,u)\,\mathrm{d}x+\frac{\mu }{q} \int _{\mathbb{R}^{N}} g(x) \vert u \vert ^{q}\,\mathrm{d}x \\ & \le \frac{1}{p} \Vert u \Vert ^{p} - \int _{\varOmega _{u} } F(x,u)\,\mathrm{d}x+\frac{\mu }{q} \Vert g \Vert _{q'} C_{{\frac{qq'}{q'-1}}} ^{q} \Vert u \Vert ^{q} \\ & \le \frac{1}{p} \Vert u \Vert ^{p} - \varepsilon ^{2}_{1} (d _{2}-d_{3}) \Vert u \Vert ^{p} +\frac{\mu }{q} \Vert g \Vert _{q'} C_{{\frac{qq'}{q'-1}}} ^{q} \Vert u \Vert ^{q}. \end{aligned}

Note that we can take sufficiently large $$d_{2}$$ (and μ sufficiently small) such that

$$\max_{u\in Y_{k}, \Vert u \Vert =r_{k}} J_{\nu }(u)< 0, \quad\forall k\in \mathbb{N}, \text{ if } \Vert u \Vert =r_{k}< \rho _{k} \text{ small enough}.$$

This completes the proof. □

From Lemmas 3.13.3, we see (Z1)–(Z3) of Lemma 2.2 hold. Therefore, there exist $$\nu _{n}\to 1$$, $$u(\nu _{n})\in Y_{n}$$ such that

$$J'_{\nu _{n}}|_{Y_{n}}\bigl(u(\nu _{n}) \bigr)=0,\qquad J_{\nu _{n}}\bigl(u(\nu _{n})\bigr)\to c_{k} \in \bigl[d_{k}(2),b_{k}(1)\bigr],\quad \text{as } n \to \infty.$$
(17)

For convenience, we denote $$u_{n}=u(\nu _{n})$$ for all $$n\in \mathbb{N}$$.

### Lemma 3.4

Suppose that all the assumptions of Theorem 1.1 hold. Then the sequence $$\{u_{n}\}$$ is bounded in E.

### Proof

Note that $$J_{\nu _{n}}(u(\nu _{n}))$$ is bounded, and we have

\begin{aligned} c+1 \ge{} &J_{\nu _{n}}(u_{n})- \frac{1}{p}\bigl\langle J_{\nu _{n}}'(u_{n}),u _{n}\bigr\rangle \\ ={}&\frac{1}{p}\nu _{n} \int _{\mathbb{R}^{N}} f(x,u_{n})u _{n}\,\mathrm{d}x -\nu _{n} \int _{\mathbb{R}^{N}} F(x,u_{n})\,\mathrm{d}x \\ &{}+\frac{ \mu }{q} \int _{\mathbb{R}^{N}} g(x) \vert u_{n} \vert ^{q}\,\mathrm{d}x - \frac{\mu }{p} \int _{\mathbb{R}^{N}} g(x) \vert u_{n} \vert ^{q}\,\mathrm{d}x \\ \ge{}& \int _{\mathbb{R} ^{N}} \mathscr{{F}}(x,u_{n}) \,\mathrm{d}x. \end{aligned}
(18)

We will argue by contradiction. If $$\|u_{n}\|$$ is unbounded in E, we assume that $$\|u_{n}\|\to \infty$$. Put $$v_{n}= \frac{u_{n}}{\|u_{n}\|}$$, and then $$\|v_{n}\|=1$$. Passing to a subsequence, there exists $$v\in E$$ such that $$v_{n}\rightharpoonup v$$ weakly in E, $$v_{n}\to v$$ strongly in $$L^{r}(\mathbb{R}^{N})$$ with $$r\in [p,p_{\alpha }^{*})$$, $$v_{n}(x)\to v(x)$$ for a.e. $$x\in \mathbb{R}^{N}$$. For $$0\le a< b$$, let $$\varOmega _{n}(a,b)=\{x\in \mathbb{R}^{N}:a\le |u_{n}(x)|< b\}$$. Next we consider two cases.

Case 1: Suppose $$v=0$$.

Then $$v_{n}\to 0$$ $$\text{ in }L^{r}(\mathbb{R}^{N})\text{ with } r \in [p,p_{\alpha }^{*})$$, and $$v_{n}(x)\to 0\text{ for a.e. }x\in \mathbb{R}^{N}$$. Let $$r_{0}$$ be as in (f4), and from (3) we have

\begin{aligned} \int _{\varOmega _{n}(0,r_{0})}\frac{{f}(x,u_{n})u_{n}}{ \Vert u_{n} \Vert ^{p}}\,\mathrm{d}x &= \int _{\varOmega _{n}(0,r_{0})}\frac{{f}(x,u_{n})u_{n}}{ \vert u_{n} \vert ^{p}} \vert v_{n} \vert ^{p}\,\mathrm{d}x \\ &\le \bigl(\varepsilon +c_{\varepsilon }r_{0}^{s-p}\bigr) \int _{\varOmega _{n}(0,r_{0})} \vert v_{n} \vert ^{p}\,\mathrm{d}x \\ & \le \bigl(\varepsilon +c _{\varepsilon }r_{0}^{s-p}\bigr) \int _{\mathbb{R}^{N}} \vert v_{n} \vert ^{p}\,\mathrm{d}x \to 0. \end{aligned}
(19)

From (f4), we know $$\tau >\frac{p_{\alpha }^{*}}{p_{\alpha }^{*}-p}$$. Thus, if we set $$\tau '=\tau /(\tau -1)$$, then $$p\tau '\in (p,p_{ \alpha }^{*})$$. From the Hölder inequality and (18), we obtain

\begin{aligned} \int _{\varOmega _{n}(r_{0},\infty )}\frac{{f}(x,u_{n})u_{n}}{ \Vert u_{n} \Vert ^{p}}\,\mathrm{d}x & = \int _{\varOmega _{n}(r_{0},\infty )}\frac{{f}(x,u_{n})u_{n}}{ \vert u _{n} \vert ^{p}} \vert v_{n} \vert ^{p}\,\mathrm{d}x \\ & \le \biggl( \int _{\varOmega _{n}(r_{0},\infty )} \biggl(\frac{{f}(x,u _{n})u_{n}}{ \vert u_{n} \vert ^{p}} \biggr)^{\tau }\,\mathrm{d}x \biggr)^{\frac{1}{\tau }} \biggl( \int _{\varOmega _{n}(r_{0},\infty )} \vert v _{n} \vert ^{p\tau '} \,\mathrm{d}x \biggr)^{\frac{1}{\tau '}} \\ & \le \biggl( \int _{\varOmega _{n}(r_{0}, \infty )}\frac{ \vert {f}(x,u_{n}) \vert ^{\tau }}{ \vert u_{n} \vert ^{(p-1)\tau }}\,\mathrm{d}x \biggr)^{\frac{1}{\tau }} \biggl( \int _{\varOmega _{n}(r_{0},\infty )} \vert v _{n} \vert ^{p\tau '} \,\mathrm{d}x \biggr)^{\frac{1}{\tau '}} \\ & \le \biggl( \int _{\varOmega _{n}(r_{0},\infty )}d_{1}{\mathscr{F}}(x,u) \,\mathrm{d}x \biggr)^{\frac{1}{\tau }} \biggl( \int _{\varOmega _{n}(r_{0},\infty )} \vert v _{n} \vert ^{p\tau '} \,\mathrm{d}x \biggr)^{\frac{1}{\tau '}} \\ & \le \bigl[d_{1}(c+1)\bigr]^{\frac{1}{\tau }} \biggl( \int _{\mathbb{R}^{3}} \vert v_{n} \vert ^{p\tau '} \,\mathrm{d}x \biggr)^{\frac{1}{ \tau '}}\to 0. \end{aligned}
(20)

Combining (19) and (20), we have

$$\int _{\mathbb{R}^{N}}\frac{{f}(x,u_{n})u_{n}}{ \Vert u_{n} \Vert ^{p}}\,\mathrm{d}x= \int _{\varOmega _{n}(0,r_{0})}\frac{{f}(x,u_{n})u_{n}}{ \Vert u_{n} \Vert ^{p}}\,\mathrm{d}x + \int _{\varOmega _{n}(r_{0},\infty )}\frac{{f}(x,u_{n})u_{n}}{ \Vert u_{n} \Vert ^{p}}\,\mathrm{d}x\to 0.$$
(21)

On the other hand, note that $$\nu _{n}\to 1$$, from (5) and (g) we have

\begin{aligned} 1 =\frac{ \Vert u_{n} \Vert ^{p}}{ \Vert u_{n} \Vert ^{p}} &=\frac{\langle J_{\nu _{n}}'(u _{n}),u_{n}\rangle }{ \Vert u_{n} \Vert ^{p}}+\frac{\nu _{n}}{ \Vert u_{n} \Vert ^{p}} \int _{\mathbb{R}^{N}}{f}(x,u_{n})u_{n}\,\mathrm{d}x- \frac{\mu }{ \Vert u_{n} \Vert ^{p}} \int _{\mathbb{R}^{N}} g(x) \vert u_{n} \vert ^{q}\,\mathrm{d}x \\ & \le \frac{\langle J_{\nu _{n}}'(u_{n}),u_{n}\rangle }{ \Vert u_{n} \Vert ^{p}}+\frac{\nu _{n}}{ \Vert u _{n} \Vert ^{p}} \int _{\mathbb{R}^{N}}{f}(x,u_{n})u_{n}\,\mathrm{d}x+ \frac{\mu C _{\frac{qq'}{q'-1}}^{q}}{ \Vert u_{n} \Vert ^{p}} \Vert g \Vert _{q'} \Vert u_{n} \Vert ^{q} \\ & \le \limsup_{n\to \infty } \biggl[\frac{\langle J_{\nu _{n}}'(u_{n}),u _{n}\rangle }{ \Vert u_{n} \Vert ^{p}}+ \frac{\nu _{n}}{ \Vert u_{n} \Vert ^{p}} \int _{\mathbb{R}^{N}}{f}(x,u_{n})u_{n}\,\mathrm{d}x+ \frac{ \Vert u_{n} \Vert ^{q}}{ \Vert u _{n} \Vert ^{p}}\mu \Vert g \Vert _{q'}C_{\frac{qq'}{q'-1}}^{q} \biggr] \\ & \le \limsup_{n\to \infty } \frac{\nu _{n}}{ \Vert u_{n} \Vert ^{p}} \int _{\mathbb{R}^{N}}{f}(x,u_{n})u_{n}\,\mathrm{d}x, \end{aligned}

Case 2: Suppose $$v\neq0$$.

Set $$A=\{x\in \mathbb{R}^{N}: v(x)\neq0\}$$ and $$\text{meas}(A)>0$$. For $$x\in A$$, we have $$\lim_{n\to \infty }|u_{n}(x)|=\infty$$. Hence $$A\subset \varOmega _{n}(r_{0},\infty )$$ for large n. From (3) and (f3), note the nonnegativity of $${f}(x,u)u$$, Fatou’s lemma enables us to obtain

\begin{aligned} 0={}&\lim_{n\to \infty }\frac{o(1)}{ \Vert u_{n} \Vert ^{p}}=\lim _{n\to \infty }\frac{ \langle J_{\nu _{n}}'(u_{n}),u_{n}\rangle }{ \Vert u_{n} \Vert ^{p}} \\ ={}& \lim_{n\to \infty } \biggl[\frac{ \Vert u_{n} \Vert ^{p}}{ \Vert u_{n} \Vert ^{p}}+\frac{ \mu }{ \Vert u_{n} \Vert ^{p}} \int _{\mathbb{R}^{N}} g(x) \vert u_{n} \vert ^{q}\,\mathrm{d}x-\frac{ \nu _{n}}{ \Vert u_{n} \Vert ^{p}} \int _{\mathbb{R}^{N}}{f}(x,u_{n})u_{n}\,\mathrm{d}x \biggr] \\ \le{}& 1+ \lim_{n\to \infty } \biggl[\frac{ \Vert u_{n} \Vert ^{q}}{ \Vert u_{n} \Vert ^{p}}\mu \Vert g \Vert _{q'}C_{\frac{qq'}{q'-1}}^{q}- \int _{ \varOmega _{n}(0,r_{0})}\frac{ {f}(x,u_{n})u_{n}}{ \Vert u_{n} \Vert ^{p}}\,\mathrm{d}x \\ &{}- \int _{ \varOmega _{n}(r_{0}, \infty )}\frac{{f}(x,u_{n})u_{n}}{ \vert u_{n} \vert ^{p}} \vert v_{n} \vert ^{p}\,\mathrm{d}x \biggr] \\ \le{}& 1+ \limsup_{n\to \infty } \int _{ \varOmega _{n}(0,r_{0})}\frac{ {f}(x,u_{n})u_{n}}{ \Vert u_{n} \Vert ^{p}}\,\mathrm{d}x-\liminf _{n\to \infty } \int _{ \varOmega _{n}(r_{0},\infty )} \frac{{f}(x,u_{n})u_{n}}{ \vert u_{n} \vert ^{p}} \vert v_{n} \vert ^{p}\,\mathrm{d}x \\ \le{} &1+ \limsup_{n\to \infty }\frac{\varepsilon r_{0}^{p}+c_{ \varepsilon }r_{0}^{s}}{ \Vert u_{n} \Vert ^{p}}\cdot \text{meas} \bigl( \varOmega _{n}(0,r _{0})\bigr) \\ &{}-\liminf _{n\to \infty } \int _{ \varOmega _{n}(r_{0},\infty )}\frac{ {f}(x,u_{n})u_{n}}{ \vert u_{n} \vert ^{p}}\bigl[\chi _{\varOmega _{n}(r_{0},\infty )}(x) \bigr] \vert v _{n} \vert ^{p}\,\mathrm{d}x \\ \le{}& 1 - \int _{ \varOmega _{n}(r_{0},\infty )}\liminf_{n\to \infty }\frac{ {f}(x,u_{n})u_{n}}{ \vert u_{n} \vert ^{p}} \bigl[\chi _{\varOmega _{n}(r_{0},\infty )}(x)\bigr] \vert v _{n} \vert ^{p}\,\mathrm{d}x\to -\infty. \end{aligned}

Thus $$\{u_{n}\}_{n\in \mathbb{N}}$$ is bounded in E. This completes the proof. □

### Lemma 3.5

Suppose that all the assumptions of Theorem 1.1 hold. For some $$\varLambda >0$$, the sequence $$\{u_{n}\}$$ possesses a strong convergent subsequence in E.

### Proof

From Lemma 3.4, the sequence $$\{u_{n}\}_{n\in \mathbb{N}}$$ is bounded in E. Then there exists $$u\in E$$ such that $$u_{n}\rightharpoonup u$$ weakly in E, $$u_{n}\rightarrow u$$ strongly in $$L^{r}(\mathbb{R}^{N})$$ for $$r\in [p,p_{\alpha }^{*})$$ and $$u_{n}(x)\rightarrow u(x)$$ for a.e. $$x \in \mathbb{R}^{N}$$ after passing to a subsequence if necessary. Next, we prove two claims.

Claim 1. $$\langle J_{\nu _{n}}'(u_{n}-u),u_{n}-u\rangle =o(1)$$ as $$n\to \infty$$.

Let $$w_{n}=u_{n}-u$$. Then $$w_{n}\rightharpoonup 0$$ weakly in E, $$w_{n}\rightarrow 0$$ strongly in $$L^{r}(\mathbb{R}^{N})$$ for $$r\in [p,p_{\alpha }^{*})$$, and $$w_{n}(x)\rightarrow 0$$ for a.e. $$x \in \mathbb{R}^{N}$$ after passing to a subsequence. Recall that $$u_{n}\rightharpoonup u$$ weakly in E, we have $$\|w_{n}\|=\|u_{n}\|- \|u\|+o(1)$$, and from (7) we only need to show

$$\int _{\mathbb{R}^{N}} f(x,w_{n})w_{n}\,\mathrm{d}x=o(1) \quad\text{and}\quad \int _{\mathbb{R}^{N}} g(x) \vert w_{n} \vert ^{q}\,\mathrm{d}x=o(1), \quad\text{as } n \to \infty.$$

In fact, from (3) we have

\begin{aligned} &\biggl\vert \int _{\mathbb{R}^{N}} f(x,w_{n})w_{n}\,\mathrm{d}x \biggr\vert \le \int _{\mathbb{R}^{N}} \bigl\vert f(x,w_{n}) \bigr\vert \vert w_{n} \vert \,\mathrm{d}x\le \varepsilon \int _{\mathbb{R}^{N}} \vert w_{n} \vert ^{p} \,\mathrm{d}x+c_{\varepsilon } \int _{\mathbb{R}^{N}} \vert w_{n} \vert ^{s} \,\mathrm{d}x \to 0,\\ &\quad \text{as } n\to \infty \text{ with } s\in \bigl[p,p_{\alpha }^{*}\bigr), \end{aligned}

and

$$\int _{\mathbb{R}^{N}} g(x) \vert w_{n} \vert ^{q}\,\mathrm{d}x \le \Vert g \Vert _{q'} \Vert w_{n} \Vert _{\frac{qq'}{q'-1}}^{q} \to 0, \quad\text{as } n\to \infty \text{ with } \frac{qq'}{q'-1}\in \bigl[p,p_{\alpha }^{*}\bigr).$$

Claim 2. There is $$M>0$$ such that

$$\int _{\mathbb{R}^{N}} \mathscr{{F}}(x,w_{n})\,\mathrm{d}x\le M.$$

From Lemma A.1 of , there exists $$\sigma (x)\in L^{r}( \mathbb{R}^{N})$$ with $$r\in [p,p_{\alpha }^{*})$$ such that

$$\bigl\vert u_{n}(x) \bigr\vert \le \sigma (x),\qquad \bigl\vert u(x) \bigr\vert \le \sigma (x)\quad \text{for } x \in \mathbb{R}^{N}, n\in \mathbb{N}.$$
(22)

Note that $$w_{n}=u_{n}-u$$, by (3), (4), and (22) we have

\begin{aligned} \int _{\mathbb{R}^{N}} \mathscr{{F}}(x,w_{n})\,\mathrm{d}x &= \int _{\mathbb{R} ^{N}} \biggl(\frac{1}{p}{f}(x,w_{n})w_{n}-{F}(x,w_{n}) \biggr)\,\mathrm{d}x \\ & \le \int _{\mathbb{R}^{N}} \biggl(\frac{2 \varepsilon }{p} \vert w_{n} \vert ^{p}+\frac{c _{\varepsilon }(p+s)}{ps} \vert w_{n} \vert ^{s} \biggr) \,\mathrm{d}x \\ & \le \int _{\mathbb{R}^{N}} \biggl(\frac{2^{p+1}\varepsilon }{p} \sigma _{1} ^{p}(x)+\frac{2^{s} c_{\varepsilon }(p+s)}{ps}\sigma _{2}^{s}(x) \biggr)\,\mathrm{d}x \\ & \le {M}, \end{aligned}

where $${M}>0$$, $$\sigma _{1}\in L^{p}(\mathbb{R}^{N}), \sigma _{2}\in L ^{s}(\mathbb{R}^{N})$$ with $$s\in (p,p_{\alpha }^{*})$$.

Now, we prove that the sequence $$\{u_{n}\}_{n\in \mathbb{N}}$$ has a convergent subsequence. Note $${V}(x)< b$$ on a set of finite measure and $$w_{n}\rightarrow 0$$ strongly in $$L^{r}(\mathbb{R}^{N})$$, $$r\in [p,p_{\alpha }^{*})$$, and we have

$$\Vert w_{n} \Vert _{p}^{p}= \int _{\mathbb{R}^{N}} \vert w_{n} \vert ^{p}\,\mathrm{d}x\le \frac{1}{ \lambda b} \int _{{V}\ge b} \lambda {V}(x) \vert w_{n} \vert ^{p} \,\mathrm{d}x + \int _{{V}< b} \vert w_{n} \vert ^{p} \,\mathrm{d}x\le \frac{1}{\lambda b} \Vert w_{n} \Vert ^{p}+o(1).$$

Combining this and the Hölder inequality, for $$s=\frac{p\tau }{ \tau -1}\in [p,p_{\alpha }^{*})$$, fixed $$\nu \in (s,p_{\alpha }^{*})$$, and we have

\begin{aligned} \Vert w_{n} \Vert _{s}^{s} &= \int _{\mathbb{R}^{N}} \vert w_{n} \vert ^{s}\,\mathrm{d}x \\ &= \int _{\mathbb{R}^{N}} \vert w_{n} \vert ^{\frac{p(\nu -s)}{\nu -p}} \vert w_{n} \vert ^{s-\frac{p( \nu -s)}{\nu -p}}\,\mathrm{d}x \\ & \le \biggl( \int _{\mathbb{R}^{N}} \vert w_{n} \vert ^{\frac{p( \nu -s)}{\nu -p}\frac{\nu -p}{\nu -s}}\,\mathrm{d}x \biggr)^{\frac{\nu -s}{\nu -p}} \biggl( \int _{\mathbb{R}^{N}} \vert w_{n} \vert ^{(s-\frac{p( \nu -s)}{\nu -p})\frac{\nu -p}{s-p}}\,\mathrm{d}x \biggr)^{ \frac{s-p}{\nu -p}} \\ & = \biggl( \int _{\mathbb{R}^{N}} \vert w_{n} \vert ^{p}\,\mathrm{d}x \biggr)^{\frac{\nu -s}{\nu -p}} \biggl( \int _{\mathbb{R}^{N}} \vert w_{n} \vert ^{ \nu }\,\mathrm{d}x \biggr)^{\frac{s-p}{\nu -p}} \\ &\le \biggl(\frac{1}{ \lambda b} \biggr)^{\frac{\nu -s}{\nu -p}}C_{\nu }^{\frac{\nu (s-p)}{ \nu -p}} \Vert w_{n} \Vert ^{\frac{p(\nu -s)}{\nu -p}} \Vert w_{n} \Vert ^{\frac{\nu (s-p)}{ \nu -p}} \\ & = \biggl(\frac{1}{\lambda b} \biggr)^{\frac{\nu -s}{ \nu -p}}C_{\nu }^{\frac{\nu (s-p)}{\nu -p}} \Vert w_{n} \Vert ^{s} \quad\text{for } C_{\nu }>0. \end{aligned}

From (f1), for any $$\varepsilon >0$$, there exists $$\delta =\delta ( \varepsilon )>0$$ such that $$|{f}(x,u)|\le \varepsilon |u|^{p-1}$$ for $$x\in \mathbb{R}^{N}$$ and $$|u|\le \delta$$. Moreover, (f4) is also satisfied for some suitable δ. Therefore, we have

$$\int _{ \vert w_{n} \vert \le \delta } {f}(x,w_{n})w_{n} \,\mathrm{d}x\le \varepsilon \int _{ \vert w_{n} \vert \le \delta } \vert w_{n} \vert ^{p} \,\mathrm{d}x \le \frac{\varepsilon }{ \lambda b} \Vert w_{n} \Vert ^{p}+o(1),$$

and

\begin{aligned} \int _{ \vert w_{n} \vert \ge \delta } {f}(x,w_{n})w_{n} \,\mathrm{d}x &= \int _{ \vert w_{n} \vert \ge \delta } \frac{{f}(x,w_{n})w_{n}}{ \vert w_{n} \vert ^{p}} \vert w_{n} \vert ^{p} \,\mathrm{d}x \\ & \le \biggl( \int _{ \vert w_{n} \vert \ge \delta } \frac{ \vert {f}(x,w_{n}) \vert ^{\tau }}{ \vert w_{n} \vert ^{(p-1) \tau }} \,\mathrm{d}x \biggr)^{1/\tau } \biggl( \int _{ \vert w_{n} \vert \ge \delta } \vert w_{n} \vert ^{\frac{p \tau }{\tau -1}} \,\mathrm{d}x \biggr)^{(\tau -1)/\tau } \\ & \le \biggl( \int _{ \vert w _{n} \vert \ge \delta } d_{1}\mathscr{{F}}(x,u) \,\mathrm{d}x \biggr)^{1/\tau } \Vert w_{n} \Vert _{s}^{p} \\ & \le (d_{1}{M})^{1/\tau } \biggl(\frac{1}{\lambda b} \biggr)^{\frac{p(\nu -s)}{s(\nu -p)}}C_{ \nu }^{\frac{p\nu (s-p)}{s(\nu -p)}} \Vert w_{n} \Vert ^{p}+o(1). \end{aligned}

Consequently, we have

\begin{aligned} o(1) &=\bigl\langle J_{\nu _{n}}'(w_{n}),w_{n} \bigr\rangle = \Vert w_{n} \Vert ^{p}+\mu \int _{\mathbb{R}^{N}}g(x) \vert w_{n} \vert ^{q} \,\mathrm{d}x -\nu _{n} \int _{\mathbb{R} ^{N}}{f}(x,w_{n})w_{n}\,\mathrm{d}x \\ & \ge \Vert w_{n} \Vert ^{p} - 2 \int _{\mathbb{R}^{N}}{f}(x,w_{n})w_{n}\,\mathrm{d}x \\ & \ge \biggl[1-\frac{2 \varepsilon }{\lambda b}-2(d_{1}{M})^{1/\tau } \biggl( \frac{1}{\lambda b} \biggr)^{\frac{p(\nu -s)}{s(\nu -p)}}C_{\nu }^{\frac{p\nu (s-p)}{s( \nu -p)}} \biggr] \Vert w_{n} \Vert ^{p}+o(1). \end{aligned}

Thus there exists $$\varLambda >0$$ such that $$w_{n}\to 0$$ in E when $$\lambda >\varLambda$$. This implies that $$u_{n}\to u$$ in E. This completes the proof. □

### Proof of Theorem 1.1

From the last assertion of Lemma 2.2, we know that $$J=J_{1}$$ has infinitely many nontrivial critical points. Therefore, (1) possesses infinitely many small negative-energy solutions. This completes the proof. □

## References

1. 1.

Ge, Z., Ou, C.: Chaos in a fractional order modified Duffing system. Chaos Solitons Fractals 34(2), 262–291 (2007)

2. 2.

Podlubny, I.: Fractional Differential Equations. Academic Press, San Diego (1999)

3. 3.

Zhou, Y.: Basic Theory of Fractional Differential Equations. World Scientific, Singapore (2014)

4. 4.

Kilbas, A.A., Srivastava, H.M., Trujillo, J.J.: Theory and Applications of Fractional Differential Equations. North-Holland Mathematics Studies, vol. 204. Elsevier, Amsterdam (2006)

5. 5.

Zuo, M., Hao, X., Liu, L., Cui, Y.: Existence results for impulsive fractional integro-differential equation of mixed type with constant coefficient and antiperiodic boundary conditions. Bound. Value Probl. 2017, Article ID 161 (2017)

6. 6.

Wu, J., Zhang, X., Liu, L., Wu, Y., Cui, Y.: The convergence analysis and error estimation for unique solution of a p-Laplacian fractional differential equation with singular decreasing nonlinearity. Bound. Value Probl. 2018, Article ID 82 (2018)

7. 7.

Hao, X., Wang, H., Liu, L., Cui, Y.: Positive solutions for a system of nonlinear fractional nonlocal boundary value problems with parameters and p-Laplacian operator. Bound. Value Probl. 2017, Article ID 182 (2017)

8. 8.

Cui, Y.: Uniqueness of solution for boundary value problems for fractional differential equations. Appl. Math. Lett. 51, 48–54 (2016)

9. 9.

Zou, Y., He, G.: On the uniqueness of solutions for a class of fractional differential equations. Appl. Math. Lett. 74, 68–73 (2017)

10. 10.

Chen, C., Xu, J., O’Regan, D., Fu, Z.: Positive solutions for a system of semipositone fractional difference boundary value problems. J. Funct. Spaces 2018, Article ID 6835028 (2018)

11. 11.

Zhang, K.: On a sign-changing solution for some fractional differential equations. Bound. Value Probl. 2017, Article ID 59 (2017)

12. 12.

Guo, Y.: Nontrivial solutions for boundary-value problems of nonlinear fractional differential equations. Bull. Korean Math. Soc. 47(1), 81–87 (2010)

13. 13.

Bai, Z., Zhang, Y.: Solvability of fractional three-point boundary value problems with nonlinear growth. Appl. Math. Comput. 218(5), 1719–1725 (2011)

14. 14.

Wang, Y., Liu, L., Zhang, X., Wu, Y.: Positive solutions of an abstract fractional semipositone differential system model for bioprocesses of HIV infection. Appl. Math. Comput. 258, 312–324 (2015)

15. 15.

He, J., Zhang, X., Liu, L., Wu, Y., Cui, Y.: Existence and asymptotic analysis of positive solutions for a singular fractional differential equation with nonlocal boundary conditions. Bound. Value Probl. 2018, Article ID 189 (2018)

16. 16.

Zhang, X., Shao, Z., Zhong, Q., Zhao, Z.: Triple positive solutions for semipositone fractional differential equations m-point boundary value problems with singularities and p-q-order derivatives. Nonlinear Anal., Model. Control 23(6), 889–903 (2018)

17. 17.

Mao, J., Zhao, Z., Wang, C.: The exact iterative solution of fractional differential equation with nonlocal boundary value conditions. J. Funct. Spaces 2018, Article ID 8346398 (2018)

18. 18.

Song, Q., Bai, Z.: Positive solutions of fractional differential equations involving the Riemann–Stieltjes integral boundary condition. Adv. Differ. Equ. 2018, Article ID 183 (2018)

19. 19.

Sun, Q., Meng, S., Cui, Y.: Existence results for fractional order differential equation with nonlocal Erdélyi–Kober and generalized Riemann–Liouville type integral boundary conditions. Adv. Differ. Equ. 2018, Article ID 243 (2018)

20. 20.

Bai, Z., Chen, Y.Q., Lian, H., Sun, S.: On the existence of blow up solutions for a class of fractional differential equations. Fract. Calc. Appl. Anal. 17(4), 1175–1187 (2014)

21. 21.

Zhang, X., Liu, L., Zou, Y.: Fixed-point theorems for systems of operator equations and their applications to the fractional differential equations. J. Funct. Spaces 2018, Article ID 7469868 (2018)

22. 22.

Zhang, X., Liu, L., Wu, Y., Zou, Y.: Existence and uniqueness of solutions for systems of fractional differential equations with Riemann–Stieltjes integral boundary condition. Adv. Differ. Equ. 2018, Article ID 204 (2018)

23. 23.

Zhang, Y.: Existence results for a coupled system of nonlinear fractional multi-point boundary value problems at resonance. J. Inequal. Appl. 2018, Article ID 198 (2018)

24. 24.

Ambrosio, V.: Multiple solutions for a fractional p-Laplacian equation with sign-changing potential. arXiv:1603.05282

25. 25.

Torres, C.: Existence and symmetry result for fractional p-Laplacian in $$\mathbb{R}^{n}$$. arXiv:1412.3392

26. 26.

Cheng, B., Tang, X.: New existence of solutions for the fractional p-Laplacian equations with sign-changing potential and nonlinearity. Mediterr. J. Math. 13(5), 3373–3387 (2016)

27. 27.

Zhang, Y., Tang, X., Zhang, J.: Existence of infinitely many solutions for fractional p-Laplacian Schrödinger–Kirchhoff type equations with sign-changing potential. Rev. R. Acad. Cienc. Exactas Fís. Nat., Ser. A Mat. 113(2), 569–586 (2019)

28. 28.

Zhang, W., Tang, X., Zhang, J.: Infinitely many radial and non-radial solutions for a fractional Schrödinger equation. Comput. Math. Appl. 71(3), 737–747 (2016)

29. 29.

Pucci, P., Xiang, M., Zhang, B.: Multiple solutions for nonhomogeneous Schrödinger–Kirchhoff type equations involving the fractional p-Laplacian in $$\mathbb{R}^{N}$$. Calc. Var. 54(3), 2785–2806 (2015)

30. 30.

Xu, J., Wei, Z., Dong, W.: Weak solutions for a fractional p-Laplacian equation with sign-changing potential. Complex Var. Elliptic Equ. 61(2), 284–296 (2016)

31. 31.

Yang, L.: Multiplicity of solutions for fractional Schrödinger equations with perturbation. Bound. Value Probl. 2015, Article ID 56 (2015)

32. 32.

Torres, C.: Non-homogeneous fractional Schrödinger equation. arXiv:1311.0708

33. 33.

Zhang, X., Liu, L., Wu, Y., Wiwatanapataphee, B.: Nontrivial solutions for a fractional advection dispersion equation in anomalous diffusion. Appl. Math. Lett. 66, 1–8 (2017)

34. 34.

Zhu, B., Liu, L., Wu, Y.: Local and global existence of mild solutions for a class of nonlinear fractional reaction-diffusion equation with delay. Appl. Math. Lett. 61, 73–79 (2016)

35. 35.

Zhu, B., Liu, L., Wu, Y.: Local and global existence of mild solutions for a class of semilinear fractional integro-differential equations. Fract. Calc. Appl. Anal. 20(6), 1338–1355 (2017)

36. 36.

Zhang, X., Liu, L., Wu, Y., Cui, Y.: New result on the critical exponent for solution of an ordinary fractional differential problem. J. Funct. Spaces 2017, Article ID 3976469 (2017)

37. 37.

Zhang, X., Liu, L., Wu, Y.: Variational structure and multiple solutions for a fractional advection-dispersion equation. Comput. Math. Appl. 68(12), 1794–1805 (2014)

38. 38.

Khoutir, S., Chen, H.: Multiple nontrivial solutions for a nonhomogeneous Schrödinger–Poisson system in $$\mathbb{R}^{3}$$. Electron. J. Qual. Theory Differ. Equ. 28, 1 (2017)

39. 39.

Shi, H., Chen, H.: Multiple positive solutions for nonhomogeneous Klein–Gordon–Maxwell equations. Appl. Math. Comput. 337, 504–513 (2018)

40. 40.

Li, L., Boucherif, A., Daoudi-Merzagui, N.: Multiple solutions for 4-superlinear Klein–Gordon–Maxwell system without odd nonlinearity. Taiwan. J. Math. 21(1), 151–165 (2017)

41. 41.

Ye, Y., Tang, C.: Existence and multiplicity of solutions for Schrödinger–Poisson equations with sign-changing potential. Calc. Var. 53(1–2), 383–411 (2015)

42. 42.

Zhang, J., Tang, X., Zhang, W.: Infinitely many solutions of quasilinear Schrödinger equation with sign-changing potential. J. Math. Anal. Appl. 420(2), 1762–1775 (2014)

43. 43.

He, J., Zhang, X., Liu, L., Wu, Y.: Existence and nonexistence of radial solutions of the Dirichlet problem for a class of general k-Hessian equations. Nonlinear Anal., Model. Control 23(4), 475–492 (2018)

44. 44.

He, X., Qian, A., Zou, W.: Existence and concentration of positive solutions for quasi-linear Schrödinger equations with critical growth. Nonlinearity 26(12), 3137–3168 (2013)

45. 45.

Liu, J., Qian, A.: Ground state solution for a Schrödinger–Poisson equation with critical growth. Nonlinear Anal., Real World Appl. 40, 428–443 (2018)

46. 46.

Liu, J., Zhao, Z.: Existence of positive solutions to a singular boundary-value problem using variational methods. Electron. J. Differ. Equ. 2014, 135 (2014)

47. 47.

Liu, J., Zhao, Z.: An application of variational methods to second-order impulsive differential equation with derivative dependence. Electron. J. Differ. Equ. 2014, 62 (2014)

48. 48.

Liu, J., Zhao, Z.: Multiple solutions for impulsive problems with non-autonomous perturbations. Appl. Math. Lett. 64, 143–149 (2017)

49. 49.

Mao, A., Chang, H.: Kirchhoff type problems in $$R^{N}$$ with radial potentials and locally Lipschitz functional. Appl. Math. Lett. 62, 49–54 (2016)

50. 50.

Mao, A., Jing, R., Luan, S., Chu, J., Kong, Y.: Some nonlocal elliptic problem involving positive parameter. Topol. Methods Nonlinear Anal. 42(1), 207–220 (2013)

51. 51.

Mao, A., Wang, W.: Nontrivial solutions of nonlocal fourth order elliptic equation of Kirchhoff type in $$\mathbb{R}^{3}$$. J. Math. Anal. Appl. 459(1), 556–563 (2018)

52. 52.

Mao, A., Yang, L., Qian, A., Luan, S.: Existence and concentration of solutions of Schrödinger–Poisson system. Appl. Math. Lett. 68, 8–12 (2017)

53. 53.

Mao, A., Zhu, X.: Existence and multiplicity results for Kirchhoff problems. Mediterr. J. Math. 14(2), 58 (2017)

54. 54.

Qian, A.: Infinitely many sign-changing solutions for a Schrödinger equation. Adv. Differ. Equ. 2011, Article ID 39 (2011)

55. 55.

Qian, A.: Sing-changing solutions for nonlinear problems with strong resonance. Electron. J. Differ. Equ. 2012, 17 (2012)

56. 56.

Shao, M., Mao, A.: Multiplicity of solutions to Schrödinger–Poisson system with concave-convex nonlinearities. Appl. Math. Lett. 83, 212–218 (2018)

57. 57.

Sun, Y., Liu, L., Wu, Y.: The existence and uniqueness of positive monotone solutions for a class of nonlinear Schrödinger equations on infinite domains. J. Comput. Appl. Math. 321, 478–486 (2017)

58. 58.

Sun, F., Liu, L., Wu, Y.: Finite time blow-up for a class of parabolic or pseudo-parabolic equations. Comput. Math. Appl. 75, 3685–3701 (2018)

59. 59.

Sun, F., Liu, L., Wu, Y.: Finite time blow-up for a thin-film equation with initial data at arbitrary energy level. J. Math. Anal. Appl. 458, 9–20 (2018)

60. 60.

Zhang, X., Jiang, J., Wu, Y., Cui, Y.: Existence and asymptotic properties of solutions for a nonlinear Schrödinger elliptic equation from geophysical fluid flows. Appl. Math. Lett. 90, 229–237 (2019)

61. 61.

Zhang, X., Liu, L., Wu, Y.: The entire large solutions for a quasilinear Schrödinger elliptic equation by the dual approach. Appl. Math. Lett. 55, 1–9 (2016)

62. 62.

Zhang, X., Liu, L., Wu, Y., Caccetta, L.: Entire large solutions for a class of Schrödinger systems with a nonlinear random operator. J. Math. Anal. Appl. 423(2), 1650–1659 (2015)

63. 63.

Zhang, X., Liu, L., Wu, Y., Cui, Y.: Entire blow-up solutions for a quasilinear p-Laplacian Schrödinger equation with a non-square diffusion term. Appl. Math. Lett. 74, 85–93 (2017)

64. 64.

Zhang, X., Liu, L., Wu, Y., Cui, Y.: The existence and nonexistence of entire large solutions for a quasilinear Schrödinger elliptic system by dual approach. J. Math. Anal. Appl. 464(2), 1089–1106 (2018)

65. 65.

Yue, Y., Tian, Y., Bai, Z.: Infinitely many nonnegative solutions for a fractional differential inclusion with oscillatory potential. Appl. Math. Lett. 88, 64–72 (2019)

66. 66.

Zhang, X., Liu, L., Wu, Y., Cui, Y.: Existence of infinitely solutions for a modified nonlinear Schrödinger equation via dual approach. Electron. J. Differ. Equ. 2018, 147 (2018)

67. 67.

Di Nezza, E., Palatucci, G., Valdinoci, E.: Hitchhiker’s guide to the fractional Sobolev spaces. Bull. Sci. Math. 136(5), 521–573 (2012)

68. 68.

Zhao, J.: Structure theory of Banach spaces. Wuhan University Press (1991)

69. 69.

Zou, W.: Variant fountain theorems and their applications. Manuscr. Math. 104(3), 343–358 (2001)

70. 70.

Willem, M.: Minimax Theorems. Birkhäuser, Boston (1996)

Not applicable.

Not applicable.

## Funding

Research supported by the National Natural Science Foundation of China (Grant No. 11601048), Natural Science Foundation of Chongqing (Grant No. cstc2016jcyjA0181), Natural Science Foundation of Chongqing Normal University (Grant No. 16XYY24), and Natural Science Foundation of Shandong Province (Grant No. ZR2018MA009 and ZR2015AM014).

## Author information

Authors

### Contributions

The authors contributed equally to this paper. The authors read and approved the final manuscript.

### Corresponding author

Correspondence to Keyu Zhang.

## Ethics declarations

Not applicable.

### Competing interests

The authors declare that they have no competing interests.

Not applicable.

Not applicable.

### Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

## Rights and permissions 